Hello everyone,
I was hoping for some definite direction in my endless quest to fix my =
OK, but be warned you might not like my answers :-).
C64. I don't have a lot of free time to work on
this. So it is ongoing =
in many ways. Plus I am cautious about doing things that I know little =
about. Which can be good especially when you get people that give you =
conflicting advice. I do find a lot of helpful people on IRC, but you =
have to sort out what they are all telling you. I won't want to go and =
start buying all sorts of stuff and equipment without knowing why or =
then finding that I should have bought Y, when I bought X.=20
I know a little. Bits and pieces here and there, but nothing that will =
give me confidence to just dive in.
This is my quest to learn how to repair my C64:
I started with my original C64 which I bought in 1987. I had left it in =
a garage for two years and then tried to use it and it didn't work.=20
First I was told that I should clean my C64 with regular dish detergent =
and a hair dryer. I was told that this would solve many problems. Sounds =
unsafe, but I guess I will try it.
Then I was told that I should buy another C64 since it isn't worth =
repairing them, since they are so plentifully available and cheap. So =
now I have many non-working ones.
Err, yes... Been therem, done that (but not with C64s...).
Whether something is worth repairing is up to you. It may well not make
financial sense to repair a C64, given how commn they are and how cheap
they tenet to me, but the fiannacial rason is not the only reason for
doing something (in fact to me it's rarely a reason at all). If you want
otlearn how to repair things, it's not a bad machine to start on. And you
will learn a lot. You might even come to ejoy it. THose seem to me to be
reasons why it is worth repariign a C64.
Much of what follows is not going to be C64-specific. That;s not
suprising, IMHO almost all computers -- darn it, all electronic devices
-- are diagnosed and repaired in much the ame way,
There are 2 parts ot the repair. First you must find out what is wrong,
second you put it right. Notice the order there. It is, IMHO, very
foolish to start trying to reapri things, replace parts, etc until you
have a lcear idea as to what is wrong amd what therefore needs replacing.
Since your questions are in the opposite order, I will anseer them like
that, though.
Then I need to replace chips that are bad, so I need
to know how to =
solder and de-solder. Which kind of device to get ? There are different =
wattages and if you do it wrong then you burn up your boards (as a =
friend of mine did with more soldering experience did). Do I get a =
My view (having strugled with them) is that the
non-tempearture-controlled hobbyist irons, things like the 25W ones that
I guess Radio Shack sells, are not worth trying to use. You will have
lots of problems and you will get frustrated fast.
A tempaerature-cotnroled iron is the thing to go for. I use a Weller TCP,
about 50W. A non-controlled 50W iron would get hot enough to damage the
board and components, bnt this has a thermostat built-in to keep the
temperature constant(ish). And being a 50W iron it has enoguh power to
heat up large bits of metal, like PCB ground planes, screenign cans, etc.
Others swear by electronic-controlled irons, still others by RF-heated
irons, etc. They are nice, but very expensive.
combination desolder sucker ? Or a little squeeze one
? Or a push and =
suck stick ? Do I get a soldering station ? A braider ? Too many =
different choices and combinations. I prefer something that will work =
and not damage my boards, and for desoldering, something that won't give =
me repetitive stress injury.
Unless you are sondiering buying a professioanl desoldering station (and
I doubt it), but one of those little spring-loaded pumps. Make sure you
can get repalcement (ptfe) tips for it, you will need them from time to
time.
You will ahve to practice using it. A couple of quick tips. Firstly, if
you are sure a component is bad, cut off the pins, remvoe the body of the
componet, and despdler the pins one at a time. Often you can just melt
the sodler and yank the pin out with pliers. To claer the holes in the
PCB (sometimes needed anyway), it's easiest ot melt the soder with the
iron on one side of the PCB and such from the other side.
If A pin won't desolder cleanly, resodler it (use soem new solder) and
try again. Often that does it.
As regards RSI. I once upgraded the memory board in an HP Integral. The
board I had was half-populakted with RAM chips, but it had (of course)
been wave-soldered at the factory, so all the spare holes were blocked. I
had to clar out 256 holes (16 RAM chips at 16 pins each) plus those for
decoupling capacitors etc, My hand ached a bit afterwards, but no harm
done...
Then I'm told to get a diagnostic cartridge for
C64, which works well =
expect when the PLA chip is bad and there you can't see video.
Then I should get a diagnostic harness, which works better, but again =
you need video.
Then I find a Diagnose 64 cartridge which tells you which chips are bad =
very simply with LED lights, but they are hard to find. I'm borrowing =
one right now from a friend and figuring out how to use now.
Then I'm told to get a multimeter, how do I use ? Which one to get ? I =
get one, then I'm told there is a better that could have been gotten for =
a little more money.=20
Then I'm told to get a logic probe. Which one ? Again, how to use ?
Then I'm told a logic probe is not as good as an oscilloscope. Which one =
? How to use ? Then I'm told that I don't need an oscilloscope.
Then I'm told to go to Ray Carlsen's site and that will have everything =
I need.
http://personalpages.tds.net/~rcarlsen/cbm.html. Which is very =
nice, but I'm a beginner and I don't need just a bunch of schematics and =
reference material. I need step by step method which explains which =
tools, techniques etc. that I need to do.
Then I'm told that going to the Rob Clarke and Bil Herd workshop would =
give me everything that I need to know. It is good information, but not =
hands on.
OK... Let's take a step back.
The wway I repair things -- the way I learnt to repair things many years
ago -- goes like this :
1) Work out what the thing _should_ be doing
2) Find out what it actually is doing
3) Compare the two
4) Figure out what could case the observed behaviour to differ from the
expected behaviour, that's the fault. If necessary do more tests to find
out exactly what is wrong.
Now let's look at some of those steps. The scheamtic sare useful to work
out wwhat it should be doing. Somebody with experience can look at a
schemaitc and have a good idea as to what the signals, voltages, etc
should be on most points in the circuit. Yes, that is a big step for a
beginner, but most service manaulas were not written fro beginners. When
you have more experience you will come to realise that often all you
really need is the schematic. Of coruse sometimes other have been
helpful and noted down foltages (e.g. power supply otuputs voltages),
signals 9e.g. master clock signals, video timing signals) so you don't
have to work them allout for yourself.
The next problem is that you cna't useuflly detect electricity yourself.
So you need measuring insturments to do that. These will tell you what
hte circuit is actually doing, step (2) of my procedure. There are many
bits of test equipment aviaalble, some general-purpose (used with jus
abotu all electrical/electornic devices), some very specialised. Let's
stick to the general-purpose ones.
A multimeter measures voltage, current or resistnace (if you are not
familiar with those terms then you really need to start by going back a
few stanges and learnign some basic electroncis before you start fixing
computers). It's by far the most used test instrument. You se it to check
power supply voltages (a very common problem), to test swithces,
connectors, etc. It's not so useful for fiding logic fualts.
An oscilloscope (oftne shortened to 'scope) is a device that displays a
graph of votlage (vertial) agaisnt time (horizontal) (OK, I know there
are many otehr thigns it can be used for, but let's keep things simple
for now). The simpler 'scopes ahve no intenral storage and are really
only useful on repetitive waveforms. For example the AC ouptut form a
mains transformer, the signals in a swithcing regulator (but take _great_
care if you start workign on these!), video timing signals, etc. More
advanced/modern 'scopes have storage. You can get them to record voltae
agaisn time for pwehaps 10us and then display the grpah for evermore. So
you could look at the continuosuly clangin signals on a computer bus.
a logic probe is a farily simple insrument that displays on LEDs wheter a
signal is a logic 1, a logic 0 or changing. Althoguh it is simple, it's
suprisingly useful. FOr example, if a comuputer is running a program, all
the data bus lines are changing as the instructions are red by the
processor. So touchign the logic probe onto each of them in turn should
show that on the probe's LEDs. If one signal is always high or low you
know you have a fault there.
An instruemnt you haven;'t mentioned and which I find vey useful is a
logic analuser. This is cross between a logic probe and a storage 'scope.
It samples nad records a number of logic signals and displays them. OK,
it doesn't record the actuall voltages, only whether they are 0 or 1, but
it will orten record faster than a cheap storage 'scope and it will
record more signals (32 channel logic analusers are commmon, 32 channel
storage 'scopes are not).
Some years ago HP made a pocket-sized logic analyser that they called an
'advanded logic probe'. It went under the name LogicDart. It was as easy
to sue as a lotgic probe, but told you rather more about the devic eunder
test. They are not easy to find now, but I find mine (bought when it just
came out) very useful.
I don;t find things like diagnostic programs or cartridges to be useful.
The point is, they are running on a piece of defective hardware so (a)
there will always be faults they can't find (a diagnostic program is not
goign to be able ot find PSU problems in a C64, for example) nad they
might well give misleading reuslts because the hardware is
malfunctioning. They have their uses for finding problems in
non-essential areas fo the machine (for exampel if a C64 ahs sound
problems, you might have a program to put the right vaulse in the SID
chip to prosduce a known output), but not for intiial troubleshooting.
-tony